Corporate interests are generally non-local resource users specialized in the extraction of commercial timber, charcoal or fuelwood. In general they gain access to forests with the goal of extracting resources for sale in regional markets, major urban centers, or for export. The primary objectives of resource extraction specialists include profit maximization and minimizing the transactions costs of conducting business. In addition, resource extraction specialists may have an interest in gaining short or long term property rights of forested land as a mechanism for ensuring a long-run supply of forest products to markets. Obtaining long-term property rights may negate the need for repeated negotiations between government authorities and other stakeholder groups that have an interest in gaining access to the resource.
Resource extraction specialists are likely to have a high level of technical capacity for the extraction of valuable forest resources. Many resource extraction specialists may be powerful actors that are well connected to both central government officials and local elites. Resource extraction specialists are also likely to be relatively wealthy and able to mobilize capital and sources of credit when necessary. Resource extraction specialists may have limited knowledge of local forest conditions and political hierarchies, particularly if they have traditionally dealt with central authorities regarding the acquisition of permits and quotas allowing them to harvest forest products.
Because resource extraction specialists are motivated by profit margins and keeping the transactions costs of conducting business as low as possible, they are unlikely to favor decentralization reforms that increase the number of actors they are required to negotiate rights of access to forest resources with. In addition, under decentralized governance structures non-local resource users including resource extraction specialists and pastoralists that may have previously negotiated access and use rights with centralized government agencies find themselves negotiating with a new set of actors that likely have different incentives for the management and conservation of forests. McCarthy (2004) found that actors engaged in extracting timber in Kalimantan encountered volatile and shifting socio-legal configurations in which they had to negotiate rights of access to forest resources. Where they had previously dealt with only central government bureaucrats, they now have to negotiate access to forest resources at every level of government. The uncertainty associated with conducting business in the decentralized environment added an additional level of risk to timber extraction.
4.1.5 Other Factors that Affect Outcomes: Biophysical and Economic Incentives
Individuals’ livelihoods and forest condition, our outcome variables, are also influenced by other factors separate from the design and implementation of institutions. These include biophysical factors and economic incentives, which must be taken into account when exploring the role of decentralization . Biophysical factors such as elevation, slope, soils, rainfall and temperature, shape the abundance and type of forest resources (Tucker, Randolph and Castellanos, 2007). Agrawal and Chhatre (2006) caution that if these factors are not considered in explaining forest condition, there is a danger of overestimating the influence of institutional and/or other variables on forest condition. In their study of determinants of forest condition in decentralized forests in India, they find biophysical factors to overwhelm the significance of economic pressures, institutions, demographic pressures and gender relations.
Biophysical factors also include the quality and quantity of the resource units, their predictability, as well as their spatial extent (Gibson, Ostrom and McKean, 2000). These attributes in turn affect the kinds and magnitude of benefits that individuals can obtain from the forests, as well as their incentives for sustainable harvesting and protection. For example, rugged terrain surrounds the El Picacho forest in Mexico (Tucker, 2004), providing protection against the large-scale illegal logging that has deforested part of neighboring San Marco forest. Similarly, despite clear and well-defined boundaries of government forest reserves in Uganda’s Mpigi District, the forests are large, with long boundaries and hence difficult to effectively monitor. A poorly resourced forest authority is incapable of equipping and motivating its officials; hence resource users can easily circumvent rules, and harvest indiscriminately (Gombya-Ssembajjwe and Banana, 2000). By influencing the productivity of a resource, or the ability to access and harvest it, or the willingness of groups and individuals to engage in conserving it, biophysical properties do indeed affect, if not determine, forest condition.
Economic incentives are critical if forest adjacent individuals and groups are to engage in ecological protection and restoration activities (Baland and Platteau , 1996). Incomes and benefits can be generated in various ways. Under different variants of decentralization, central or municipal governments can transfer a given proportion of revenues to lower governance levels, or even directly to communities. In Uganda for example, 40% of gross revenue collected from forest products is repatriated to local areas where it have been generated (Nsita, 2005b). In Cameroon, close to 25% of the selling price of mature trees in cash is paid to communities (Oyono, 2005a).
Various governments in Latin America provide similar incentives to different actors (Larson, 2005b). The general point of undertaking various revenue-sharing programs is to encourage communities to sustainably manage forest resources, and sometimes to even reward them. In Uganda’s Bushenyi District, for example, motivated by increases in timber prices and increasing revenues from legally harvested timber, local timber harvesting associations track down illegal timber dealers and report them to Forest Department officials (Nsita, 2005b). It is worth noting that many governments have devolved forests that are degraded (Colfer , 2005 in Larson, 2005b; see Banana et al, 2007 for Uganda) and with little scope of generating benefits to locals. Indeed the degradation may in some cases be so severe as to preclude the effectiveness and profitability of management.
Increasingly, scholars and practitioners are realizing the importance of market access for various forest products, including non timber products. Market access can help diversify local communities’ income base, contributing to their livelihoods. However, mass commercialization of forest products can threaten resource sustainability. The production of Matsusake mushrooms in China dropped from 530 tons to 272 tons between 1995 and 2000, even as they generated 50-80% of household cash incomes and close to 30% of country revenues (Xu and Ribot, 2005a). In Cameroon, communities are tolerant of accelerated and intensive exploitation of timber (Oyono, 2005a), since returns from timber are much higher than transfers obtained through various taxes and fees. In short, while marketization can contribute to sustainable livelihoods, it also risks unleashing a race to capture rents, which, if unchecked can have negative impacts on resource condition and on benefit distribution as those with the most resources and better technologies are more likely extract more. Economic incentives are also evident in the threat of (or actual) conversion of forests into monoculture for the production of biofuels as countries seek to diversify their energy sources in attempts to mitigate climate change. Mabira forest in Uganda provides an example of a high biodiversity tropical forest that is under considerable threat of conversion to sugarcane plantation. There are similar concerns as rapidly developing countries, with high timber demands create powerful incentives for rapid, indiscriminate extraction of timber in countries with high natural forest cover. Beyond threatening sustainability, such concessions may exclude local communities from forests.
5.1 Livelihood Outcomes
Decentralization is not an end in itself. The goal of the reforms is, in most cases, to improve local welfare. But on what criteria is this to be judged, and how is this to be measured?
Improvements in local livelihoods would be an important starting point for this. Conventional measures of consumption, income, or asset accumulation can provide indicators of livelihood improvements. The data can be drawn from household (or intra-household) surveys, focus groups, or key informant interviews. In practice, however, there are particular challenges to accurately measuring such changes in the context of decentralization.
The first challenge is that, as in any impact assessment, establishing a counterfactual or comparator group is essential to show how decentralization reforms have affected the outcome indicators. Randomly assigning individuals or communities to a treatment (decentralized) or control (status quo) group, while convincing to reviewers, is generally not feasibility for decentralization studies. Indeed, because decentralization reforms are usually passed at the central government level, there would not be sufficient variability between units in the same country, except that implementation is generally not uniform. This allows us to compare between sites even within one country, if there is variability in implementation.
Even with a comparator group, we need before and after measures to identify the change in conditions. Unfortunately, not enough areas have baseline data to do this. In this context, the IFRI data sets are particularly valuable because they have collected multiple rounds of data for many sites, with a consistent core set of information for each. However, because most IFRI data are collected at the unit of the forest itself, rather than individual household users, the data are stronger for measuring changes in the resource condition than local livelihoods.
Assessing changes in forest-based livelihoods presents further challenges because of the heterogeneous products and uses that people make of forests. Men and women, local and non-local people will use the forest differently. Timber and products that are marketed are likely to be easier to measure than non-timber forest products, especially those for home consumption. How should access to medical plants, which may only be used occasionally, be valued? Should only local livelihood changes be counted, or also those of other people who live farther away but use the forest occasionally? More inclusive approaches will give a more accurate picture of the positive or negative changes, as well as of the distributional impacts, but they are also more expensive.
Related to this is the issue of timing and period of recall. In general, the shorter the period of recall, the more accurate will be measures of regular types of income or consumption. But some forest products may only be used once a year or once in several years, and short periods of recall may miss these uses .
5.2 Sustainability Outcomes
Forests are important sources of wealth and power, suggesting that shifts in property rights have implications for those ceding and gaining control (Kaimowitz and Ribot 2002). Decentralization reforms are expected to allow poor people to make their assets more productive via the mechanism of government policies that support local organization and decision-making, and the participation of local forest users in markets for forest products.
When local resource users are involved in the formation of rules regarding the use of forests they are expected to respect and enforce rules to a higher standard then when they are not involved in rule formulation (Gibson, McKean and Ostrom 2000). It is also argued that local resource users have a much greater knowledge base and will therefore be better equipped to identify and prioritize environmental problems (Larson 2002). In addition it is argued that both the financial and transaction costs of sustainable management should be lower when local resource users are integrated into the management of forests (Kaimowitz et al. 1998)
Empirical evidence of the link between decentralization reforms and forest sustainability is highly variable both within and between countries that have enacted reforms, and has been explicitly addressed in relatively few studies. Andersson and Gibson (2006) found that local governments had highly variables responses to decentralization reforms, which in turn significantly influenced deforestation rates in Bolivia. They note that local governments generally chose one of three paths with respect to forest management in their municipalities: do a lot; do a little; or ignore new powers. In cases where local governments demonstrated a high degree of interest and investment in forestry related activities rates of deforestation were lower. Oyono (2005) reports similarly mixed effects for the case of Cameroon’s extensive forest sector reform citing high rates of forest loss in some of the communities that received forest governance powers. He identifies overlapping regulations, lack of clarity regarding new forest rights associated with reforms, and an unwillingness of the central government to step intervene to avert cases of excessive forest loss as the primary reasons for high rates of forest loss in some areas. A recent study conducted in the central region of Uganda, Banana et al. (2007) found a high degree of diversity in forest conditions influenced by the demand for forest products and crops grown in forest soils, the degree of forest degradation, and the diversity of stakeholders engaged in forest management influenced forest cover change after the forest sector reform. They cite lack of clarity among District Forest Officers, local leaders and user groups regarding who is de facto in control of forest management in this area, and drastic reductions in local government budgets for forestry related activities as contributing to forest degradation.
Two other studies are worthy of note. In a remote sensing analysis of forest cover in Indonesian Borneo, Curran et al. (2004) found that deforestation was accelerated under decentralization. They cite the responsibility of local governments for issuing logging permits as an important factor contributing to higher rates of forest cover loss. The incentive for local governments to earn large revenues from issuing logging permits presents a challenge for deforestation in decentralized systems with high value forest products. In a historical study examining deforestation in Mexico’s Yucatan Peninsula, Klepeis (2003) found that rates of deforestation were higher under centralized forest governance than under decentralized systems. The mixed findings of studies on sustainability outcomes associated with decentralization reforms point to the complexity of understanding how governance reforms affect forest sustainability. Andersson and Gibson (2006) point out that a central problem with the design of studies that address the sustainability effects of reforms is that they don’t take into account institutional, macroeconomic, and other confounding factors which are influence reform outcomes.
Evaluating the sustainability impacts of decentralization reforms is a challenge. Ideally data indicating rates of change in forest cover and forest quality from the pre-reform and post-reform periods can be compared. These data may include detailed biophysical data for a representative sample of plots throughout the area of interest, data on household perceptions of change, and/or remote sensing data at a fine enough scale to illuminate both changes in forest cover and quality. However, even with a strong research design, the need to rule out alternative explanations for forest cover change is daunting. Several possible explanations need to be ruled out including: changes in the demand or supply of forest products that are strong indicators of forest health, changes in the demand or price of food crops that are grown on land recently converted from forest; and environmental factors such as climate change that may be having negative effect on forest health.
6. Decentralization in Uganda and Kenya
6.1 Uganda’s Forest Sector Governance Reform
The Forest Sector Umbrella Programme (FSUP) initiated in 1999 was the Government of Uganda’s approach to reforming the forest sector. FSUP was a multi-donor program, led by the Government of Uganda through the Ministry of Water, Lands and Environment (MWLE), and coordinated by the Forest Sector Co-ordination Secretariat (now the Forestry Inspection Division within the MWLE). The goals of FSUP were two-fold: to create a positive, effective and sustainable policy and institutional environment for the forest sector in Uganda and, through this, to increase economic and environmental benefits from forests and trees, particularly for the poor and vulnerable (MWLE 2004). Uganda’s forest sector governance reform was part of a larger government-wide restructuring motivated by the 1995 Constitution and the 1997 Local Government Act intended to downsize the public service, rationalize government functions, and improve the effectiveness and efficiency of public service provision(MWLE 2006).
The FSUP supported a number of processes in the reform of the sector, including: a forest sector review; and the development of the Uganda Forest Policy (MWLE 2001), the National Forest Plan (MWLE 2002), and the National Forestry and Tree Planting Act (GOU 2003). A major outcome of the reform process was the abolition of the centralized Forestry Department and the creation of the decentralized District Forestry Service and the for profit parastatal National Forestry Authority. The Uganda Wildlife Authority continued to manage the15 percent of Uganda’s forests and woodlands in national parks and game reserves (Table 2).
Table 2: Forested land under different categories of ownership/management, percent
Pre 2003 Reform
Post 2003 Reform
(Central and Local forest reserves; private and customary land)
Uganda Wildlife Authority
(National Parks and Game Reserves)
District Forest Service
(private and customary forest land; Local Forest Reserves)A
National Forest Authority
(Central Forest Reserves)
Uganda Wildlife Authority
(National Parks and Game Reserves)
Tropical high forest
A. Local Forest Reserves account for less than 1% of the total forest area of Uganda. Source: Adapted from MWLE (2001), data from National Biomass Survey, 1999.
The organizational details of the three forest management authorities are characterized in Table 3. The District Forestry Service is an example of democratic decentralization to local governments. As noted in Table 2, 70 percent of Uganda’s forests are currently managed by local governments. The National Forestry Authority took over Central Forest Reserves where the majority of Uganda’s high value timber and forest biodiversity is concentrated. In addition to the development of the District Forestry Service and the National Forestry Authority, the Forestry Inspection Division (FID) was created. The FID, housed in the Ministry of Water, Lands and Environment is responsible for policy and regulation in the forestry sector overseeing the activities of both the NFA and DFS.
Table 3: Characterization of Forest Governance in Uganda1
Revenues from forest product transportation and sale permits;
Revenue from timber harvesting and charcoal burning permits
Donors (years 1-4) ; Lease of forested land or sale of individual trees for timber harvesting; Auction of illegally harvested timber; Leasing of reserve land for plantation development; Contract services (mapping; seed sales; technical advice)
Gate receipts concessions; Central government; Donors; NGOs
Local government employees
Central government employees
Level of staffing
Inadequate for mandate
Adequate for mandate
Adequate for mandate
Data sources include: author’s primary research; MWLE (2006), MWLE (2002), and UWA (2004).
Table 4: Reform Mechanisms for Catalyzing Favorable Livelihood and Sustainability Outcomes1
District Forestry Services
National Forestry Authority
Mechanisms for improving livelihoods
Small business development in the forestry sector
Increased access to forest resources for subsistence use especially wild foods, building materials and medicines
Security of land and tree tenure
Sell seed and provide technical advice to land owners that want to establish plantations
Use of appropriate technologies to enhance food security (e.g. improved stoves)
Ecotourism cash or in kind contributions to communities
Promoting tree planting and forest conservation for soil fertility
Mechanisms for improving forest management
Tree planting and agroforestry development
Management and protection of reserves
Increased awareness of the value of trees on private and customary lands
Collaborative Forest Management agreements
Improved advisory services
Investment in commercial plantations
Securing land and tree tenure
Use of appropriate technologies to minimize biomass loss during cooking, harvesting and processing
Increased awareness of ecosystem services and biodiversity
Establishing community forests
1. Source is MWLE (2002).
During the reform process several mechanisms were identified for promoting favorable livelihood and sustainability outcomes. The opportunities for achieving outcomes are different for forests on private and customary lands vs. gazetted forests in Central Forest Reserves. Similarly the incentives and constraints for achieving outcomes vary with the capacity of the organizations charged with implementing policies, and also with the incentives and constraints of local and non-local resource users. Mechanisms identified as opportunities for catalyzing favorable livelihood and sustainability outcomes are summarized above in Table 4.
A parastatal is an organization which has some political authority and serves the state indirectly. While clearly identified as a parastatal in the policy documents, NFA also has a revenue generating function
The NFA Start-up Fund was supported by the European Union (EU), and the governments of Norway (NORAD) and the United Kingdom (DFID) who provided funds to meet initial capital investment and start-up operational costs. The EU provided financial assistance through the Forest Resources Management and Conservation Program (FRMCP)(MWLE 2006).
6.2. Decentralization in Kenya
Unlike Uganda, Bolivia and Mexico, decentralization of the forestry sector comes to Kenya rather late in the game, with enabling legislation only having been passed in 2005. Prior to this, Kenya was guided by the forest policy of 1957, which was revised in 1968, and then again in 1994. The main legislation is the Forest Act Cap 385 of 1962 that has been twice revised in 1982 and 1992. It was drafted in support of the 1957 policy and covered a broad range of activities from the gazettement/degazettement of forests and Nature Reserves, licensing of use, prohibitions of certain activities and imposition of penalties etc. Subsidiary regulations covered the rights of forest adjacent communities to utilize specified resources in specific ways. This Act had several crucial shortcomings. It covered only gazetted forest reserves, did not provide sufficient safeguards against forest excisions, provided only use rights to a narrow set of resources for communities, and did not recognize the importance of forests for environmental conservation.
Forest management in Kenya cannot however be entirely separated from broader administrative efforts. When Kenya attained self rule from the British colonial government in 1963, customary authorities were the main tools of administrative management (Mamdani, 1996) but these structures were taken over by the central government. In 1983, the government attempted to decentralize its activities through the creation of the District Focus for Rural Development (Ngethe, 1998). The decentralization of administrative functions of government from headquarters in the capital of Nairobi to the districts also necessitated the then Forest Department to decentralize its functions from the headquarters, provinces and divisions to the districts. During this process, the Forest Department created district forest offices across the country, including in Nairobi. The Forest Department also posted officers at Divisional level, which is one administrative level below the district level. In general the Forest Department had representation at district and divisional level in 72 districts and 180 administrative divisions. The core functions of the Forest Department included management of natural forest and water catchment areas, development and management of industrial forest plantations, promotion of farm forestry, forest protection, conservation and management of dry-land forests.
Decision-making power in the Forest Department was heavy at the top. There was little feedback along the hierarchy save for the writing of monthly, quarterly or annual reports. Fiscal responsibility was also unidirectional and the revenue collected by the Forest Department was handed to the central government and rarely ploughed back to improve the forests. The establishment of the lower level offices also did not include the transfer of financial resources from the Forest Department headquarter to the district levels. Forest officers were under resourced and lacked the basic capacities to enforce rules in forests within their jurisdiction. Because of this, there was lack of accountability in the lower level offices. By failing to lay ground for the establishment of competent local level institutions the Forest Department created a vacuum which was used by the provincial administration to thwart the proposed good management of the forest sector. Forest Department was restricted to the active management of plantations, law enforcement to control illegal extraction, licensing of forest product extraction such as firewood and fire protection, plus control of problem animals by the Kenya Wildlife Service. Forest management objectives were preservationist, excluding local resource users from decision making and forest management, with minimal and stringent provisions for subsistence extraction and use of forest products. In general, the Forest Department wielded tremendous power and authority over forest resources, with no accountability to local communities living adjacent to forest areas.
In spite of the strict protectionist strategy adopted by forest managers, the destruction and degradation of Kenya’s forest resources has formed a persistent theme in the evaluation of the natural resources sector. These have been captured in broad literatures, from newspapers (Matiru, 1999) through to technical reports by local and international agencies (Wass, 1995; KFMP, 1996). Authorized and unauthorized excisions are the prime source of forest loss. Between 1995 - 1999, a total of 44,502.77 ha. were officially de-gazetted and excised (Matiru, 1999). Between 1972-1980 the forests shrank at an annual average rate of 2% (Ochanda et al, 1981), and later at an average of 3700-5000 ha per year (Wass, 1995). The industrial plantations have faced a similar plight. While the loss of forest cover through excisions for settlement and agriculture as well as uncontrolled exploitation of forest products is thought largely to be the result of a rapid increase in population, it must be noted that the Forest Department has had little capacity to implement its extensive mandate. The Forest Department has been severely understaffed, just as the institutional framework at the time severely limited opportunities for management innovations. But even within the reality of a highly centralized forest management regime, one exceptional innovation stands out. In the mid-1990s, a special decree from the forest authorities permitted the piloting of community forest management in select forest areas in the country. This somewhat decentralized management sought to directly involve communities in forest management and monitoring, while sharing benefits and revenues from forest resources. This innovation has proven to be a valuable experience in the subsequent decentralization reform.